Frangible Roof Theory - BS EN 14015 Appendix K

[sajk14]

Is anyone familiar with Appendix K of the tank design standard BS EN 14015 regarding design rules for frangible roof to shell joints? I have found a few queries in this section of the code and would like anyones opinon on the subject........

For some reason it only addresses frangibility for unanchored tanks....Does this imply that only unanchored tanks can be considered frangible!?

I am not sure if the formulas for unstiffened lengths for the upper and lower courses lcyl and lcylb are correct in section K4 Design Rules. I have checked the formulas for a variety of tank sizes and design parameters and the values of the unstiffened lengths always come out as greater than the actual tank height!?....It's also not clear what units should be used i.e. is DF and ecyl in metres or millimetres?

The same query about using correct units m or mm applies to the remaining formulas in Section K4 for calculating roof failure pressure and bottom failure pressure.

I appreciate any thoughts and feedback.

 

[mariog123]

I can only speculate EN intentions.
Probably they tried to consider a minimum longitudinal distance from the junction edge to an adjacent structural
discontinuity as 2.5*sqrt(R*t), somehow similar to ASME VIII Div2.
Such formula has to be consistent in units, so R[mm] and t[mm] give the distance in mm.

In case we prefer the diameter instead radius, R=D/2, so the distance should be 2.5*sqrt(D*t/2).

With D[mm] and t[mm], you get the distance in [mm], so in [m] it should be [2.5/1000]*sqrt(D*t/2)

However, someone smarter than me considered the proper result as 2.5*sqrt[D*t/(2*1000)]=2.5*sqrt[D*t/2000]
My guess, of course!

 

[mariog123]

"The same query about using correct units m or mm applies to the remaining formulas in Section K4 for calculating roof failure pressure and bottom failure pressure."

Well, here I think it is clear that EN standard considers dimensionless factors as e/D, X, Y, etc, so all geometric dimensions must be entered in the same units, in this case [mm], exactly as Table K.1 suggests.

In this case, the millibar units (for results after applying pressure formulas) would be hidden in alfa1, beta0, beta 1 and beta2 values.

Never used that section of EN and I have no idea what model/scientific paper they considered to calculate the design roof and bottom failure pressure. In fact I have no idea which is the failure mechanism they considered and from where are coming up such formulas. Just curious about the values calculated vs. API650 values, can you give some preliminary conclusions of your calculations?

 

[mariog123]

"For some reason it only addresses frangibility for unanchored tanks....Does this imply that only unanchored tanks can be considered frangible!?"

IMO, it's about the validity of the theory considered. Their theory assumes, in line with ASME VIII Div2, that there is a minimum distance of junction edge to an adjacent structural discontinuity; here the condition is to avoid elements that could add stiffness to junctions, because the theory cannot count them properly...
So they imposed lcylb as unstiffened length for the lower course and the anchorage disturb the theory.

PS. Please excuse me for posting three times trying to replay, it was an fragmented weekend!
After all it's not more than "thoughts and feedback" on the subject, as you asked for.

My best regards.

 

[sajk14]

Mariog123,

Thank you for taking time to look through my post and for your feedback. I will try to obtain a copy of ASME VIII Div2 regarding the similar formula/theory.

When I get a bit of time I will try to carry out some frangibility calcs for same tanks under both API 650 and EN 14015 rules for a comparison. I will let you know my findings.

With regards to EN 14015 frangible roof theory. Generally I have found that when you use the formulas in section K4 ( using consistent units of mm as you mention above ), the tanks almost always come out as NOT frangible.

This is generally due to 2 reasons........

1) When you calculate the required unstiffened shell 'lcyl' and 'lcylb' dimensions to satisfy frangibility criteria, the value most of the time comes out as greater than the actual tank shell height.

2) When calculating the roof failure pressure and bottom failure pressure using the stated formulas ( again all units in mm ), generally the roof failure pressure is substantially greater than the bottom failure pressure ( i.e. the shell to bottom joint will fail before the roof to shell junction ).
EN 14015 states that the bottom failure pressure should be at least twice the value of the roof failure pressure which is very difficult to achieve using this theory.

In most cases when the roof is not frangible under EN 14015 i would typically recommend anchoring the tank and specify the use of emergency/explosion vents.

Does anyone one else in the forum have any thoughts on this or can shed some light on the subject?

Best regards

 

[mariog123]

I would add that:
- ASME VIII Div2 hasn't a similar theory, however 2.5*sqrt(R*t) is mentioned in few sections as the distance junction- structural discontinuity.
- 'lcyl' and 'lcylb' are wrong written in EN 14015 Appendix K, and you would correct them as 2.5*sqrt(D*t/2) [mm]- at least, this formula is similar with others considered in Codes;
- according to API 650/ API pb 937, the target of the frangible roof-to-shell joint should be that the yielding of the top compression ring will occur before uplift of the bottom-base of the shell. IMO, API 650 5.10.2.6 (5) would be interpreted that at roof failure (corresponding to pf pressure and Fy stress in compression ring), the vertical force Vf (roof force can be separated into vertical and radial components) satisfies Vf< DLS. So the criteria considered are different between API and EN. According to your calculations, there are cases when API 650 declares "frangible" a tank and EN 14015 says that tank is "clearly non-frangible", due to reason no 2?

Thanks a lot.

 

[sajk14]

Thanks Mariog123,

So what approach would you use as to whether or not these formulas are applicable or not for a tank with anchors?

i.e. all formulas etc shown in Section K refer to unanchored tanks only. Therefore if a tank requires anchoring we cannot apply section K theory and possibly revert to another code for frangibility criteria for anchored tanks..........say API 650 for e.g?

 

[mariog123]

Dear skn4,

As I tried to explain, API 650 considers a frangibility theory as the yielding of the top compression ring will occur before uplift of the bottom-base of the shell. As a consequence, it is not important the tank is anchored or not. In my opinion API 650 criteria means that the shall-bottom is not loaded when the roof is collapsing. That's why it appears to be a conservative criteria in practice, but it is a clear, engineering recognizable, criteria.

EN 14015 Appendix K tries to considers different scenario and calculates roof-shell failure pressure and bottom-shell failure pressure. The failure pressures theory is not available for the reader and, exactly as you said, the criteria is "the bottom failure pressure should be at least twice the value of the roof failure pressure".
One problem is that anchorage and/or supplementary stiffening on top course are not complying with the theory driving the calculations of the failure pressures; that's why Appendix K requires for 'lcyl' and 'lcylb' dimensions without any additional stiffening. What is happening when we have anchorage? Pb (bottom-shell failure pressure) will be (substantial?) more than those calculated and we cannot say how much it is, that's all.

I'm afraid your questions haven't answers... A Code must be applied exactly "as is". But exactly as you observed, EN 14015 Appendix K failed to cover all cases (anchored tanks are not subject to their theory) and, to be worse off, they failed to give valid formulas for 'lcyl' and 'lcylb'.
Frankly speaking, this is almost a rule; EN codes are full of theory (intended to be more advanced) but quite difficult to be applied in practice and full of typing (and not only typing) mistakes.

My best regards.

 

[mariog123]

"API 650 criteria means that the shall-bottom is not loaded when the roof is collapsing" was intend to be "API 650 frangibility criteria means that the longitudinal stress in the junction shell-bottom is zero (no uplift) when the roof is collapsing".
I guess fixing a more reasonable limit of the the longitudinal stress would lead to a more reasonable criteria of frangibility, but this discussion is out of API 650 interpretation.

 

[sajk14]

Thanks mariog123 for your insight and thoughts on the matter, it has been good to discuss this issue with someone.

Its a shame no one else on the forum was able to get involved in this discussion ( Although this may be because the EN 14015 code is probably not used by the majority of the members on here ).

Thanks and Best Regards

 

[mariog123]

It was a discussion about an appendix of an European Code, limited in usage and not transparent in requirements, where units are not clearly specified in Nomenclature section, where some formulas are wrong written and you would be in doubt that the rest of them are valid....
I would say it is a shame that they sell such Codes and I cannot blame people never going to get into discussing of such [...].

 

[IFRs]

I was following this discussion but had nothing substantive to offer.